Semantic Landmark Particle Filter for Robot Localisation in Vineyards

TL;DR

This paper introduces a Semantic Landmark Particle Filter that enhances robot localisation in vineyards by integrating semantic structural cues, significantly reducing errors caused by repetitive row patterns.

Contribution

The novel SLPF method combines semantic trunk detections with LiDAR data and GNSS priors to improve localisation accuracy in repetitive vineyard environments.

Findings

Reduces Absolute Pose Error by up to 65% compared to baselines

Improves row correctness from 0.67 to 0.73

Decreases mean cross-track error from 1.40 m to 1.26 m

Abstract

Reliable localisation in vineyards is hindered by row-level perceptual aliasing: parallel crop rows produce nearly identical LiDAR observations, causing geometry-only and vision-based SLAM systems to converge towards incorrect corridors, particularly during headland transitions. We present a Semantic Landmark Particle Filter (SLPF) that integrates trunk and pole landmark detections with 2D LiDAR within a probabilistic localisation framework. Detected trunks are converted into semantic walls, forming structural row boundaries embedded in the measurement model to improve discrimination between adjacent rows. GNSS is incorporated as a lightweight prior that stabilises localisation when semantic observations are sparse. Field experiments in a 10-row vineyard demonstrate consistent improvements over geometry-only (AMCL), vision-based (RTAB-Map), and GNSS baselines. Compared to AMCL, SLPF reduces Absolute Pose Error by 22% and 65% across two traversal directions; relative to a NoisyGNSS baseline, APE decreases by 65% and 61%. Row correctness improves from 0.67 to 0.73, while mean cross-track error decreases from 1.40 m to 1.26 m. These results show that embedding row-level structural semantics within the measurement model enables robust localisation in highly repetitive outdoor agricultural environments.